High throughput genomic platforms have revealed a profound effect of age on gene expression and gene methylation levels. For example, our integromic analysis of multiple publicly available data sets have identified hundreds of age related genes that can be organized into network modules. These gene sets along with genes and pathways known from the vast literature on aging and longevity provide valuable candidates in allelic association studies of the health and well-being of older Americans. The rich longitudinal data set and the genome wide scans of DNA samples from the Health and Retirement Study (HRS) provide a unique resource for evaluating the phenotypic effects of genetic variants underlying aging related genes, gene sets and for developing multivariable models that include behavioral, psychosocial, and genetic factors. This proposal aims to apply systems biologic and systems genetic methods for identifying gene sets based on existing gene expression, gene methylation, and multiple large scale genome wide association studies (GWAS). Single nucleotide polymorphisms (SNPs) underlying these genes will then be related to health, cognitive, behavioral, and economic phenotypes using multivariable regression-, machine learning-, and weighted network analysis approaches. An iterative process between SNP and phenotype selection facilitates the definition of clinically and economically important phenotypes under genetic control (reverse phenotyping). The proposal will not only elucidate the genetic and molecular underpinnings of retirement relevant phenotypes (cognitive functioning, psychosocial factors, and health related expenditures) but also lead to a phenotypic annotation of aging related genes and pathways.